Non-IR Wing Leveler

[rclinks2002]

Hello everyone,
I am having an issue locating a feasible wing leveler (for UAV) that doesnt use IR as a reference. (FMA Co pilot will not work) Does anyone have any ideas? Thanks

Ben Mitchell

[Gary Warner]

When you say wings-leveler... What's the real control issue you are trying to resolve? Heading? Bank? Rate of turn? Image stabilizing, etc.?

This topic has been batted around here quite a bit. Many approaches have been taken, but the 'best' resolution revolves around the 'real' goal (control issue) for a wing-leveler.

Some of the sensors have been magnetometers, accelerometers, gyros (piezo and mechanical), thermal (you've already ruled this one out), lasers (see attitude indicator of glass-cockpit GV), GPS, and/or combinations of the above, etc. Other ideas might fail my memory right now.

Please share more about what the real application is and what the real control issue is. If we can focus on the desired effects of a wing-leveler, we'd be in a better position to share useful help.

Gary
--

[rclinks2002]

Gary,
Thanks for your reply. By means of leveler I mean a way to keep an airplane wing level without any control inputs from the ground. What I am planning on using it for is to have an autopilot that only controls the rudder, as you know, when you try to control an airplane (non glider) with only rudder it wants to drop the inside wing. I need the wing leveler to prevent the airplane from dropping that wing, and allow a flat turn with the rudder. Basically it needs to do exactly what the FMA Direct Co pilot does but without Infrared Reference.....Any thoughs? Thanks

Ben Mitchell

[Gary Warner]

Umm...

I have a problem with the rudder/wing drop logic. Do I dare open this Pandora Box? ...oh, what the heck :D

When you turn on the rudder, all you are doing is changing the perceived angle-of attack (AOA) of each side of the wing in a differential mode. That is, one wing produces more lift, the other produces less. It induced a roll.

Now you want to stop the roll, right?

So, if there is no roll, how do you expect to turn?

Quick, before the flames char me to a black pile of meat... I think I understand what it is you want to do.

I think what you want is a Roll Dampener. You don't want a large heading change (or best read as 'ground track' change) just because the plane is in the effect of the rudder - flying in a skid or slip, right? But at the same time, I think it's a fair assumption that you do want to change heading (ground track) sooner or later, all be it in a controlled fashion. This will require some degree of induced roll.

My first thought is to advise you to approach the roll dampening as the focus of your project. But, in the application you describe (rudder for heading control - before I go off the deep end, let it be known now that I do not sanction rudder control for heading control in a UAV... repeat... in a UAV - other flight objectives, ok, just not UAV's) I think it would be wise to think of yaw dampening as a means of roll stabilization. After all, if the slip/skid is controlled (that is the rudder effect) then roll is controlled, to some extent.

One of the reasons it's hard to say 'just this' or 'just that' (like my last sentence) would work best is that the effects of attitude control rarely affect just one axis. Like you have recognized, rudder input (inducing a skid or slip in flight yaw direction) can and most likely will effect another axis (roll, as you have said, needs to be controlled).

Before I get into a religious discussion about flight dynamics, I want you to again ask you to reduce the objective to a more defined goal. It will really allow me to pay more attention to one aspect of flight over another. Not that only one effect can be resolved in absolute terms, but that the goals of the control issue can be 'best' resolved.

So, let me ask a few more questions, and if it helps you, you can grade the answer by degrees of importance. In other words, say this is more important then that.

1. Is this UAV to fly a GPS derived course?

2. Is this UAV to fly a goal-and-return course or need to make ground-tracking course changes?

3. Does this UAV need to keep its nose pointed to one land-based object for extended periods of time?

4. Does roll in itself (without heading change) detract from the UAV's flight goals?

5. Does the UAV use earth-based magnetic tracking (compass heading)?

6. Is the UAV expected to recover from unusual attitudes autonomously (unusual meaning over +/- 30 degree pitch or roll errors from level flight)?

7. Is the UAV to fly autonomously in extended climbs and descents?

8. Does the airframe exhibit naturally stable characteristics (Positive dihedral effect, forward CG effect, etc.)?

9. What is the degree of errors (heading/roll/yaw/track/airspeed/etc.) that are to be considered acceptable (ALL autonomous air vehicles have 'some' degree of flight errors - what are your required limits?)?

(I so want to round this post out with a #10 question, but it's late :p)

Don't get me wrong, I'm not spewing decent here, I'm just trying to understand. I don't waste my typing time to answer with long-winded responses unless I'm truly interested in the topic. Like I've said, it's been hashed here before and I have yet to see one answer fits all automatic control issues possed. I think the more we explore these kinds of issues, the better we will be ready to provide DIY electronic(s) answers to basic questions like yours.

Gary
--

[Comatose]

Gary,
Thanks for your reply. By means of leveler I mean a way to keep an airplane wing level without any control inputs from the ground. What I am planning on using it for is to have an autopilot that only controls the rudder, as you know, when you try to control an airplane (non glider) with only rudder it wants to drop the inside wing. I need the wing leveler to prevent the airplane from dropping that wing, and allow a flat turn with the rudder. Basically it needs to do exactly what the FMA Direct Co pilot does but without Infrared Reference.....Any thoughs? Thanks

Ben Mitchell

Polyhedral? Freeflight guys have been making airplanes that are controlled by only rudder and power for a long, long time now.

A combination of accelerometer and gyroscope can be used as a wing leveler (or more correctly a bank control system, why limit yourself to flat when you can pick any reasonable orientation.) but if your real goal is a cheap "goes up, loiters on some approximate course then descends" an airframe with great inherent stability, a digital compass, an altimeter and a GPS will get you there a lot faster than playing with active stability control on roll and pitch.

Its also easier to gimbal a camera than it is to maintain strict attitude control, if the objective in staying flat is to keep a camera pointed directly at something.

[dalbert02]

I am curious what precludes the use of the FMA copilot. Geometry, environment, ??? It may be possible to modify the FMA system rather then seeking alternate methods of accomplishing the same goal.
-dave

[rclinks2002]

The FMA copilot if I remember right reads the difference between the grounds IR signuature and the horizons. So it would level according to temperature differences betwen the ground and the horizon...If I am not mistaken...

Ben

[dalbert02]

Kinda, it reads the signature of the ground and the sky relative to the horizon. With the Z sensor, it can even right itself from inverted. The greater the temperature difference between the ground and the sky, the more successful it is. Fly over snow and it may not provide the optimal wing leveling, fly over black asphalt and it could not get any better. Usually, grass or light brush is most common terrain flown over and it handles this just fine. That is why I was curious why it would not work in your application. What are you attempting?
-dave

[HELModels]

If the thermopiles are whacked out by Humungous ground source of IR, then maybe adding an extra set of thermopiles and then writing PIC program to filter out the outlying IR affecting the flight control thermopiles.

IOW, If there is one source of IR that is throwing the use of thermopiles out as a solution, then add an extra set of termopiles and identify the amount of effect and filter it out by software.

[JMP_blackfoot]

Back in the seventies, Maynard Hill developed an Electrostatic Autopilot, which made use of the earth's electrical field. Voltage levels within the atmosphere are smooth and horizontal. A description of atmospheric electricity can be found in the November 1972 issue of "Astronautics and Aeronautics". The voltage gradient at ground level is typically 180 Volt per meter. The system designed by Maynard uses two radioactive sources (polonium, as found in antistatic brushes) to ionise the air and provide good electrical contact. The sources are located at each wing tip, so a voltage difference between them appears if the wings are not level. It is this voltage which is used to effect any required correction of the model's attitude. A full description of the system, including schematics appeared in "Radio Control Manual no. 5" published by Model and Allied Publications Ltd in the UK.

[Mr.RC-CAM]

A useful discussion regarding Maynard's leveler is found here: http://www.rc-cam.com/forum/index.php?showtopic=740.